Cursor 0.50 update: looking slick!
We’re talking simpler pricing (finally!), a beefed-up “Max Mode,” a sneak peek at agents that work while you do other stuff, and a ton more. Let’s dive in!
Money talks!
First up: pricing changes.
Cursor heard the calls to make things less confusing, and they’ve rolled out a unified, request-based pricing model. Essentially, all your model usage now falls under this umbrella. The “Max Mode,” which gives you the full beans of their top models, now uses token-based pricing – think of it like how you pay for API access, giving you more direct control over your spend when you’re tackling those super gnarly problems.
Also, don’t worry though, the quotas on Hobby, Pro, and Business plans haven’t changed, and those handy slow requests are still included. You can track all your usage in your dashboard, which is always a plus for keeping an eye on things.
Max Mode, it’s now available for all state-of-the-art models in Cursor. When you absolutely, positively need the most context, intelligence, and tool use for a beastly coding challenge, Max Mode is your go-to. For everything else, the trusty normal mode is still there with all its familiar capabilities.
New Tab Model & Background Agents!
Now for some of the really juicy AI enhancements!
Cursor has trained a new Tab model that’s not just about suggesting code in one file – oh no, this bad boy can suggest changes across multiple files! This is a dream for refactors, chaining edits, and when you’re hopping between related bits of code. They say it also feels more natural and snappier in day-to-day use, and who doesn’t love a bit more snap? Plus, syntax highlighting for completion suggestions? Yes, please!
But they’re also rolling out an early preview of Background Agents! Imagine kicking off a bunch of AI agents to work on tasks in parallel – fixing nits, doing investigations, or even writing first drafts of medium-sized PRs – all while you keep on working on something else. These agents run in their own remote environments, and you can check their status, send follow-ups, or jump in and take over whenever you need.
Context is King, and Editing Gets an Upgrade!
We all know that with AI, context is everything. Cursor is making it easier to feed your AI assistant the right info by allowing you to use @folders to add your entire codebase into context. Just enable “Full folder contents” in settings, and you’re golden. If a folder is too massive, a little icon will let you know.
The Inline Edit (Cmd/Ctrl+K) feature also got a UI refresh and some cool new options. You can now do full file edits more easily or send a selected code block straight to an agent if you need more firepower or want to make multi-file changes from that starting point. And for those super long files the agent now has a new search & replace tool that makes edits much more efficient.
Workspaces & Chat: More Power to Your Workflow!
For those of us juggling multiple projects like a circus performer, Cursor now supports multi-root workspaces. You can bring multiple codebases into the same space, and Cursor will index them all, making them available for its AI magic. This is perfect when your frontend is in one folder and your backend in another, but you want to work on them seamlessly.
And finally, a couple of nice enhancements to the Chat feature. You can now export your chats to Markdown, which is super handy for documentation or sharing. Plus, you can duplicate chats, allowing you to explore different conversational paths with the AI without losing your original thread. It’s all about giving you more control and flexibility.
The Verdict? Cursor is Leveling UP! 🚀
Simpler pricing, more powerful Max Mode, smarter suggestions, background agents doing your bidding, better context, slicker editing, and multi-project support… Cursor is rapidly becoming your AI coding command center!
